CN-122017713-A - Multi-dimensional mapping detection method and system for transient wave recording precision of fault indicator

CN122017713ACN 122017713 ACN122017713 ACN 122017713ACN-122017713-A

Abstract

The invention relates to the technical field of electrical performance test, in particular to a multidimensional mapping detection method and a multidimensional mapping detection system for transient wave recording precision of a fault indicator, and aims to solve the problems that in traditional fault indicator transient wave recording precision detection, evaluation dimensions are single, deep physical reasons cannot be revealed, and a test signal is disjointed from an actual test signal. The method comprises the steps of constructing a parameterized virtual behavior twin body, identifying and extracting sensitive parameters, generating a dynamic precision pressure test excitation sequence, synchronously inputting the dynamic precision pressure test excitation sequence to equipment to be tested and the virtual body for double-ring track mapping to obtain track deviation characteristics, positioning specific physical parameters causing deviation and generating a diagnosis result by analyzing the mapping relation between the track deviation characteristics and the sensitive parameters, and finally updating the virtual body parameters according to the diagnosis result to generate a calibrated virtual equipment model which can predict equipment performance degradation, monitor online state and simulate behavior under specific power grid disturbance, thereby expanding the value boundary of detection work.

Inventors

HUANG BINBIN
YIN JIANKUN
QI BIN
ZHANG GUIYU

Assignees

科大智能电气技术有限公司
烟台科大正信电气有限公司

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. The multi-dimensional mapping detection method for the transient wave recording precision of the fault indicator is characterized by comprising the following steps of: S1, acquiring a principle model and a nominal parameter set of a fault indicator to be detected, and constructing a parameterized virtual behavior twin body based on the principle model and the nominal parameter set; s2, identifying and extracting sensitive parameters with the largest response change to external excitation from the parameterized virtual behavior twin; S3, generating a dynamic precision pressure test excitation sequence based on the sensitive parameters; S4, synchronously inputting a dynamic precision pressure test excitation sequence to the fault indicator to be tested and the parameterized virtual behavior twin body, and synchronously collecting a first response sequence of the fault indicator to be tested and a second response sequence of the parameterized virtual behavior twin body; S5, performing double-loop track mapping based on the first response sequence and the second response sequence to obtain track deviation characteristics; S6, obtaining a mapping relation between the track deviation characteristics and the sensitive parameters, analyzing the track deviation characteristics based on the mapping relation, positioning out target sensitive parameters causing deviation, and generating a diagnosis result containing parameter offset; and S7, updating parameters of the parameterized virtual behavior twin according to the parameter offset in the diagnosis result, and generating a calibrated virtual equipment model.
2. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 1, wherein the identifying and extracting the sensitive parameter with the largest change to the external excitation response from the parameterized virtual behavior twin comprises: setting a standard broadband test signal covering a target working frequency band in the parameterized virtual behavior twin; Applying a tiny disturbance of a preset step length to each candidate parameter in the parameterized virtual behavior twin body near a nominal value of the candidate parameter, and obtaining corresponding output response deviation; Calculating partial derivatives of the output response deviation relative to the disturbance quantity of the corresponding parameters to obtain local sensitivity coefficients of the candidate parameters; And sorting the local sensitivity coefficients of all the candidate parameters, and extracting the parameters of the preset quantity which are sorted in front or the parameters of which the local sensitivity coefficients exceed a preset threshold value as the sensitivity parameters.
3. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 1, wherein the generating the dynamic precision pressure test excitation sequence based on the sensitive parameter comprises the following steps: Classifying the sensitive parameters into a first class of parameters characterizing the high frequency response, a second class of parameters characterizing the low frequency characteristic, and a third class of parameters characterizing the frequency selectivity; Generating a first excitation subsequence for exciting high frequency response capability according to the first type of parameter; Generating a second excitation subsequence for testing low frequency stability according to the second class parameter; Generating a third excitation subsequence for assessing spectral fidelity according to the third type of parameter; and carrying out time domain combination and scheduling on the first excitation subsequence, the second excitation subsequence and the third excitation subsequence to generate a dynamic accuracy pressure test excitation sequence.
4. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 1, wherein the performing the double-loop track mapping based on the first response sequence and the second response sequence to obtain the track deviation feature comprises: extracting a multidimensional feature vector from the corresponding first response sequence and second response sequence for each excitation unit in the dynamic precision pressure test excitation sequence; Forming a first dynamic track and a second dynamic track in a feature space respectively based on the multidimensional feature vectors corresponding to all the excitation units; Calculating a measurement value of an area surrounded by the first dynamic track and the second dynamic track in a feature space to obtain a track surrounding area feature; analyzing the main direction vectors of the first dynamic track and the second dynamic track in the feature space, and calculating the included angles between the main direction vectors to obtain track trend deviation angle features; The track surrounding area features and the track trend deviation angle features are weighted and combined to generate track form difference degrees; Track morphology differences are measured as track deviation features.
5. The method according to claim 4, wherein for each excitation unit in the dynamic accuracy pressure test excitation sequence, extracting the multidimensional feature vector from the corresponding first response sequence and second response sequence comprises: performing time-frequency analysis on corresponding response fragments in the first response sequence and the second response sequence to obtain a first time-frequency distribution diagram and a second time-frequency distribution diagram; extracting frequency domain energy distribution characteristics from the first time-frequency distribution diagram and the second time-frequency distribution diagram; Performing similarity calculation and time point statistics on the response fragments to respectively obtain waveform similarity characteristics and time sequence jitter statistical characteristics; And combining the frequency domain energy distribution characteristic, the waveform similarity characteristic and the time sequence jitter statistical characteristic into a multidimensional characteristic vector.
6. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 5, wherein the steps of forming a first dynamic track and a second dynamic track in a feature space based on multidimensional feature vectors corresponding to all excitation units respectively include: constructing a high-dimensional precision state space taking waveform similarity characteristics, time sequence jitter statistical characteristics and frequency domain energy distribution characteristics as coordinate axes; mapping all multidimensional feature vectors generated by the fault indicator to be tested in response to a high-dimensional precision state space according to an excitation sequence, and connecting the multidimensional feature vectors to form a first dynamic track; And mapping all multidimensional feature vectors generated by parameterized virtual behavior twin body response to a high-dimensional precision state space according to the same excitation sequence, and connecting the multidimensional feature vectors to form a second dynamic track.
7. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 1, wherein the steps of obtaining the mapping relation between the track deviation feature and the sensitive parameter, analyzing the track deviation feature based on the mapping relation, locating the target sensitive parameter causing the deviation and generating the diagnosis result including the parameter offset comprise the following steps: inputting the track deviation characteristics into a mapping analysis model stored with previewing simulation data; Matching target sensitive parameters most relevant to the current track deviation features through a mapping analysis model; And calculating the parameter offset of the target sensitive parameter based on the quantization relation recorded in the mapping analysis model.
8. The method for detecting the multidimensional mapping of the transient wave recording precision of the fault indicator according to claim 1, wherein updating the parameters of the parameterized virtual behavior twin according to the parameter offset in the diagnosis result to generate the calibrated virtual equipment model comprises the following steps: according to the diagnosis result, replacing the nominal value of the target sensitive parameter in the parameterized virtual behavior twin body with an actual correction value containing the parameter offset; Performing compensatory fine adjustment on secondary associated parameters with electric coupling relation with target sensitive parameters by taking the actual correction value as a fixed constraint condition and running closed-loop simulation; and solidifying the updated actual correction value and the parameter set subjected to the compensatory fine adjustment into a parameterized virtual behavior twin body to complete model updating, and generating a calibrated virtual equipment model capable of representing the current physical real state of the fault indicator to be tested.
9. The method for detecting the multidimensional mapping of the transient wave recording accuracy of the fault indicator according to claim 1, further comprising, after generating the calibrated virtual device model: acquiring verification excitation independent of a dynamic accuracy pressure test excitation sequence; Inputting verification excitation into the fault indicator to be tested and the calibrated virtual equipment model at the same time, and obtaining actual response and simulated response; calculating the consistency degree between the actual response and the simulated response; Confirming the validity of the calibrated virtual equipment model according to whether the consistency degree exceeds a preset threshold value; receiving real transient wave recording data acquired by a fault indicator to be tested in field operation; Inputting the real transient recording data into the calibrated virtual equipment model to obtain a predicted response waveform; And comparing the predicted response waveform with the real transient recording data, and evaluating the accuracy stability of the fault indicator to be tested in the real running environment.
10. The utility model provides a multidimensional mapping detecting system of fault indicator transient state record precision which characterized in that includes: The virtual behavior twin construction module is used for acquiring a principle model and a nominal parameter set of the fault indicator to be detected and constructing a parameterized virtual behavior twin based on the principle model and the nominal parameter set; the sensitive parameter identification and extraction module is used for identifying and extracting sensitive parameters with the largest response change to external excitation from the parameterized virtual behavior twin; The excitation sequence generation module is used for generating a dynamic precision pressure test excitation sequence based on the sensitive parameters; the data acquisition and synchronization module is used for synchronously inputting the dynamic precision pressure test excitation sequence to the fault indicator to be detected and the parameterized virtual behavior twin body, and synchronously acquiring a first response sequence of the fault indicator to be detected and a second response sequence of the parameterized virtual behavior twin body; The track mapping analysis module is used for executing double-ring track mapping based on the first response sequence and the second response sequence to obtain track deviation characteristics; the diagnosis result generation module is used for acquiring a mapping relation between the track deviation characteristics and the sensitive parameters, analyzing the track deviation characteristics based on the mapping relation, positioning target sensitive parameters causing deviation and generating a diagnosis result containing parameter offset; And the virtual equipment model calibration module updates parameters of the parameterized virtual behavior twin according to the parameter offset in the diagnosis result to generate a calibrated virtual equipment model.

Description

Multi-dimensional mapping detection method and system for transient wave recording precision of fault indicator Technical Field The invention belongs to the technical field of electrical performance testing, and relates to a multi-dimensional mapping detection method and system for transient wave recording precision of a fault indicator. Background The fault indicator is an important monitoring device widely applied to the power distribution network, and one of the core functions is to accurately and rapidly capture and record current or voltage transient waveforms, namely transient wave recording, at the moment of fault when the power distribution network has faults such as short circuit or grounding. The accuracy of the recorded wave data directly relates to the accuracy of subsequent fault property judgment, fault point positioning and accident analysis, so that the effective detection of the transient state recorded wave accuracy of the fault indicator is important. Currently, for detecting the transient wave recording precision of the fault indicator, a technical scheme is generally adopted, wherein one or more preset transient test signals simulating typical faults are injected into the fault indicator to be detected by using a standard signal generator in a laboratory environment. And comparing the waveform data recorded by the fault indicator to be detected with the corresponding standard test signal waveform, and judging whether the recording accuracy meets the requirements of the related standard by calculating errors of the waveform data and the standard test signal waveform on one or more key indexes such as amplitude, phase and response time. However, the above prior art solutions have significant drawbacks. Firstly, the evaluation dimension is relatively single and mutually isolated, only the errors of a few discrete indexes are concerned, the comprehensive performance of the recording data in the aspects of overall waveform form fidelity, complex frequency component response accuracy, dynamic time sequence stability and the like can not be comprehensively reflected, and misjudgment is easy to occur. Secondly, when the out-of-tolerance precision is detected, the scheme can only present an error result, and cannot effectively reveal the deep physical cause of the error, so that difficulty is brought to fault diagnosis and targeted improvement of equipment. Finally, the fixed and ideal test signals are adopted for detection, the complexity and the diversity of the actual transient process in the power grid are disjointed, and the actual reference value of the detection result is limited. Disclosure of Invention In view of this, in order to solve the problems set forth in the background art, a method and a system for detecting a multidimensional mapping of transient wave recording accuracy of a fault indicator are provided. The invention provides a multi-dimensional mapping detection method for transient wave recording precision of a fault indicator, which comprises the following steps of S1, obtaining a principle model and a nominal parameter set of the fault indicator to be detected, and constructing a parameterized virtual behavior twin body based on the principle model and the nominal parameter set. S2, identifying and extracting the sensitive parameter with the largest change to the external excitation response from the parameterized virtual behavior twin. And S3, generating a dynamic precision pressure test excitation sequence based on the sensitive parameters. S4, synchronously inputting the dynamic precision pressure test excitation sequence to the fault indicator to be tested and the parameterized virtual behavior twin body, and synchronously collecting a first response sequence of the fault indicator to be tested and a second response sequence of the parameterized virtual behavior twin body. S5, performing double-loop track mapping based on the first response sequence and the second response sequence to obtain track deviation characteristics. S6, obtaining a mapping relation between the track deviation feature and the sensitive parameter, analyzing the track deviation feature based on the mapping relation, locating a target sensitive parameter causing deviation, and generating a diagnosis result containing parameter offset. And S7, updating parameters of the parameterized virtual behavior twin according to the parameter offset in the diagnosis result, and generating a calibrated virtual equipment model. The invention provides a multidimensional mapping detection system for transient wave recording precision of a fault indicator, which comprises a virtual behavior twin construction module, a parameter analysis module and a fault analysis module, wherein the virtual behavior twin construction module acquires a principle model and a nominal parameter set of the fault indicator to be detected and constructs a parameterized virtual behavior twin based on the principle model and the nominal p